Scott Morgan, Michelin

TH: So, I’d like to start out, can you describe a little bit about your role in the company and how you use prototyping in general?

SM: Sure. As you know, I am in the PE4 -TS group. Well, its got a new name now, but as you would remember it, when you worked here... Its what it is. Our goal is to create next generation products in tourism and light trucks. There it is. So, we’re typically doing work that is anywhere from three to eight years out from market development. So we’re very much a research group here as opposed to development groups which you would have like TCAR or the OE group there. So, we do the initial work starting at the concept level, which may even be up to ten years out, in which we’re taking a lot more risk, doing a lot of exploratory work. Because, we’re not close to introducing anything into the market. And then we might find, out of fifteen different solutions you try, maybe three of those might show some fruit to carry on. But that’s normal when you’re doing exploratory work in a concept development mode. Then you might take those three or four elements and you take it to the next stage that we call belise type work. And that’s probably what you were working somewhat with Damon on. Although he may have been doing some concept work.

And in that phase, you are typically three to five or six years from market launch at that point. So you’re taking those most promising ideas and you’re studying them in more depth. And you’re seeing, ”what are the risks associated with these on a more detailed level.” Because, as you can imagine, the further you are away from introducing something to the market, the more risk you can take. And the closer you get to the market, much less risk.

TH: You have to start getting that out...

SM: You gotta manage it. You never get rid of your risk, but you’re managing it to a different level. Its being more refined. So, in doing that, depending on where we are in that time frame. Before things get to launching commercially a product, we can take different levels of refinement and risk in the development of the solutions we’re pursuing. So, we could take some, for example: we might take a look at a sample in a laboratory machine, in which we’re trying to understand the characteristic of maybe its stress-strain behavior. So, we would take it and make a sample out of it. It looks nothing like a tire obviously. You’re doing testing on it, you’re trying to understand the characteristics of it. Be it material stress-strain curves, fatigue resistance, resistance to environmental attack or temperature. There’s many different types of things you could do. In a sense, I would call that a prototypes. It may not match what you’re thinking of in particular. You may be thinking of something much closer to the real object, the real tire or like an object. But I could see it answering the question a multitude of ways.

TH: Right, I would define a physical prototype as any physical model of a design or a component of the design or anything used to test a function of that design. So, I would consider your physical testing of the material some form of prototyping.

SM: Ok. So, we’ve got many examples if we go back to that level. And again, I think I mentioned on the phone, what we’re really trying to manage is resources.

If we had unlimited money and an unlimited time, why would we need prototypes?

We’d just make the end product, put all of your resources in, or just make an infinite number or study tires with real molds and real materials, processes and all of that.

But the fact is, in any product development, no matter where you are, you have to make some decisions. What can we accept in terms of refinement of our solution?

How much confidence do we have in our predicted results? That really drives our

attitude about prototypes in a sense. I can give some examples on that. That’s sort of our big picture approach to it. We’ve got to get the right information we need at a given time in the time-line. So, concept phase, we don’t need a lot of detail, we’re just trying sometimes some crazy ideas. Are they even, do they even match what we think they’re going to do? If not, then abandon them. Learn from it and move on.

The closer we get to the commercial launch again, that’s when we see some things that have really born some fruit. And now we need to say, before we put something like this out on the market, we really have to evaluate it to a large degree.

TH: So, you in your group, you’re that far out concept development and testing.

SM: Our group varies from concept development all the way to belise. I’ve probably done more work on the belise development myself. So, that would be my role, getting back to that. It is technical project leader for some of our next generation tire products coming out. So, we would be developing on the team, new sculptures for tires and tread patterns. We are introducing new materials. We’re looking at new construction of the internal components in the tire and the architecture and things like that. But in our group in particular, we’re working mostly in the summit area.

Which is sculpture design, tread compound development and design, architecture in the summit area. That’s where we in our group put most of the focus. We’ve got a sister group, PE4-TCA, which does the architecture. They do more the internal, the carcass, the endurance aspects. They’re more experts in that area. So, that’s what we do. And our deliverable, is a package of tuning rules of the new sculpture that we would hand over to the TOU’s, the development groups, so that at that point they’re closer to market launch and marketing at that point now has a much clearer view of what that product needs to be. So, whereas at the beginning of our project in belise phase or even more so far removed at the beginning of the concept project, marketing

just has a less refined view at that point, because, literally, they don’t know what the market is going to do. They have some general ideas what its going to do, but they don’t know everything at that point. So, our goal is to do exploratory work where we say, if we’re varying this performance, what’s the trade-off with other performances.

We’re trying to map that domain and understand it so we can hand over a toolbox to those development groups so that wen they get their marching orders from marketing, they can then say, ahh, lets start taking these solutions that were developed at the belise level and start combining them with the rules that the belise team gave to us in combining technologies to hit a specific point where they want to be. So, that’s kind of our role in the team. And then, Jeff, is in the same group that I’m in. He probably gave you a similar description.

TH: Yea. So, can you describe a prototype that has been really successful in your design process. Has it been really helpful or extra insightful. Something that was maybe unexpected?

SM: We’ve got so many pertinent examples. I probably need to think about what would be the best.... To talk about that. The best project... How about if I start talking about some of the different ideas about prototypes and maybe that will lead me to a good one. Ok, lets talk about about. Lets go completely upstream. New material development. We use a small prototype mixing shop here... To make com-pounds. As opposed to using a big industrial process, they would use like at one of the mixing plants. What we’re trying to evaluate is a few kgs of material. And, some of the big mixing processes are processing hundreds of kgs. The scale-up difference when you go from a small shop to a big shop, you’re getting different amounts of for example heat distribution throughout the mixture. Different amounts of work, the mechanical working of the material is different. So, therefore, the product that you’re

getting out with the same ingredients going in. The conditions in which you think are operating, really aren’t exactly the same. We know there are differences. Now, a lot of times, what we’ll do, is if we’re evaluating four of five new compounds and we only have very limited amount of material, raw material available for any one of those, we can’t go to the big shops. And we can’t get time on the machines, because they’re running wide open usually, being used for production. So, we’ll use the small shop over here. But even if we’re comparing a mix of a compound that’s operating in industry today. You might be inclined to look at that reference and test it. But we’ll actually go and remake those reference compounds in the shop as well and hopefully relatively between your different solutions you’re comparing, you’ll have, you’ll an-swer the questions that you want to anan-swer. Now, there will most likely be an offset on all of these absolute compared to what we would expect when we scale that out to a different case.

TH: So, the effects of the scaling should be similar on each of the different compounds?

SM: Well, we hope so. We hope that the relative comparison between these com-pounds is representative of what would be in the full scale process. But, in the end if you’re getting in some new raw materials, different sort of behaviors, you’re putting different amounts of energy in at the small scale than you would at the large scale.

That could transcend into different physical properties. So, its a risk that we take on that. But, its a risk that we think is manageable in most cases. And we continue to ask the questions and look at the results in the end, does it make sense or not. That would be one example of a prototype. Another one was, the material sampling that I mentioned with shear samples. You’re getting stress-strain behavior on it and you’re hoping to generate, you do generate stress-strain curves, or heat curves, or failure curves of some sort. That you can use in finite element simulation models to be

com-pared analytically to other materials, just on a curve form. That to me is another form of prototyping. Another form would be like we have here, with the rapid proto-typing. Obviously that is not a tire, but its a rapid prototyping machine that would be used to build up a print if you will, via representation of the tire sculpture. And that can be very very insightful, to go from something that is pretty good in a model like Catia, where its a 3d model where you can fly in and get into detail and look at it from perspective views, local views and global views. But sometimes actually taking and printing an object like this that you can hold, a tangible object that gives you insight that you wouldn’t normally gain just from the use of the CAD alone. In fact, we use this a lot, even to show marketing. Say, ”here are the three solutions”.

And we can build these fairly quickly. Because in CAD, you can probably make these changes in days, hours to days, and you can print these things up in maybe a couple of days, you can have one generated. To do that with an actual physical tire, months and months involved in all of that and extremely expensive. We want to use rapid prototyping like this as much as we can to answer questions.

TH: Does it provide much value from an engineering standpoint? I know, when I was working here I was sitting next to the industrial design guys and they were always making these more to show marketing and look at the aesthetics of the tire. But, as far as, the functionality and the performance, is there much you can learn from doing this?

SM: Not as much, but one thing you can do is use it to compare two direct a real product like this and get a comparison like, ”oh, we’ve actually introduced a lot more void in this area” and it maybe wasn’t as evident in the CAD model even though we could measure in there. Sometimes, physically seeing it can be worth a whole lot of information. Now, there are other things you can do. You can print

with different materials. We can print with a very rubber like materials. So, you could imagine taking a small sample, it doesn’t have to be a full crown like this, and making something for a test device. Or, if you’re using the same material, for different solutions for example. What if you were to then take those and mount them on a base and then put them through physical testing? It might be very interesting.

You could learn things like stiffness for example.

TH: So, the relative stiffness between geometries, with the same material.

SM: Exactly. Or you could even be creative. You could say, we could change material.

Materials, and see how different geometries would respond differently to different materials. We look also at the coupling of geometry with material properties as well.

You could imagine doing something like that, it becomes more physically interesting for something. We have done work in our group with molding small tread block pieces with actual rubber in just a mini mold. So that you’re getting these small tread blocks that then you can take through and put in some sort of compression/shear solicitation or maybe even something that’s physically sliding. And you’re measuring frictional coefficients and things like that. So those are excellent examples of prototyping that we’ve done. Another one that we typically use a lot, we used it from time to time in projects is carving tires. That’s a very good example of a prototype in which you start with either a slick tire that you physically mold and build it like you would any other tire, but it just doesn’t have any sculpture on it. Then you take it and they use their hot knives and different other means to remove material and can quickly come up with designs. Its rather labor intensive, but its a lot faster turn around and a lot less expensive than making physical components. So, again, ways to go faster and cheaper to get the answers that you want to get. That’s the real goal of prototyping.

Some other examples. We can actually make pieces of molds with rapid prototyping.

Such that we could mold small tread elements on them and do some other physical type testing on those as well, to understand behaviors and maybe you could simulate the process instead of just the performance. Those are other things you could do with a tire. Mold making. Mold making as I mentioned is very time consuming. Very labor intensive. Very expensive to do that. What if we could make a prototype of a full mold of the tire itself, very rapidly. That would be very interesting for us. Now, what if that mold had kind of a rough aspect to it, it wasn’t real aesthetically clean.

But, it was able to represent the areas, local areas of the sculpture that we want, the depths of the sculpture. Basically, all the major characteristics we were looking for but the fit and finish wasn’t that great on it? Or say, it was a little discretized. You could see that they used some sort of method that wasn’t as smooth as a continuous method. But you could take molded tires and run and test them very quickly. That would be interesting. We’ve done that in the past.

TH: How have you done that?

SM: That’s pretty proprietary, so I’d rather not say the details on that so much. But again, thinking creatively to come up with ways to meet the need at the moment. To go fast, and cheaper to get the answers that you want.

TH: OK.

SM: that’s really the goal here. When we make real molds for production... One thing about Michelin tires, if you compare a Michelin to anyone else, we completely stand apart aesthetically. That’s because we have a really high level of detail in our process. We are a premium product, and we want to appear as if we’re a premium product. Therefore we have many constraints. Well, if you can relax those constraints for development, why not? If you don’t have to hold to the same level of quality, that

doesn’t have any impact on performance, or extremely small impact on performance, why not relax those? So, some examples that we have... How about this? You make a tire, a physical tire with a good mold, like we would for selling to the public. And you say well, what happens if I take this tire and go buff some tread off of it? Take it down to different tread depths? You’re doing studies, that’s a prototype that you’re making there where you can go and compare a full tread depth reduced tread depth.

You didn’t make a new mold to do it. Or used another means, a precision buffing means to take rubber off. I don’t know if Jeff talked about that or not.

TH: I think I’m familiar with that process from when I was working here.

SM: That’s a prototype. That’s a definition of prototype in which you’re taking something that you can make and modifying it afterwards to create something new.

So, I do have an example of a successful one. Its this tire here. You see some of the features down here inside the grooves?

TH: Yes.

SM: That’s a complicated feature in there that’s very hard to make. We wanted to know the impact of that. What’s its impact in hydro-performance? What’s its

SM: That’s a complicated feature in there that’s very hard to make. We wanted to know the impact of that. What’s its impact in hydro-performance? What’s its